Discovery of Conformationally Restricted Human Glutaminyl Cyclase Inhibitors as Potent Anti-Alzheimer's Agents by Structure-Based Design

Article abstract of DOI:10.1021/acs.jmedchem.9b00751

Alzheimer's disease (AD) is an incurable, progressive neurodegenerative disease whose pathogenesis cannot be defined by one single element but consists of various factors; thus, there is a call for alternative approaches to tackle the multifaceted aspects of AD. Among the potential alternative targets, we aim to focus on glutaminyl cyclase (QC), which reduces the toxic pyroform of β-amyloid in the brains of AD patients. On the basis of a putative active conformation of the prototype inhibitor 1, a series of N-substituted thiourea, urea, and α-substituted amide derivatives were developed. The structure-activity relationship analyses indicated that conformationally restrained inhibitors demonstrated much improved QC inhibition in vitro compared to nonrestricted analogues, and several selected compounds demonstrated desirable therapeutic activity in an AD mouse model. The conformational analysis of a representative inhibitor indicated that the inhibitor appeared to maintain the Z-E conformation at the active site, as it is critical for its potent activity.

Full text of DOI:10.1021/acs.jmedchem.9b00751

Subscriber access provided by Nottingham Trent University
Article
Discovery of Conformationally Restricted Human Glutaminyl Cyclase
Inhibitors as Potent Anti-Alzheimer’s Agents by Structure-Based Design
Van-Hai Hoang, Van T.H. Ngo, Minghua Cui, Nguyen Van Manh, Phuong-
Thao Tran, Jihyae Ann, Heejin Ha, Hee Kim, Kwanghyun Choi, Young Ho Kim,
Hyerim Chang, Stephani Joy Y. Macalino, Jiyoun Lee, Sun Choi, and Jeewoo Lee
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00751 • Publication Date (Web): 14 Aug 2019
Downloaded from pubs.acs.org on August 14, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of Conformationally Restricted Human
Glutaminyl Cyclase Inhibitors as Potent Anti-Alzheimer’s
Agents by Structure-Based Design
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Van-Hai Hoang,†,# Van T.H. Ngo,⊥,# Minghua Cui,‡,# Nguyen Van Manh, Phuong-Thao Tran,
†
¶
†
§
§
§
§
‡
Jihyae Ann, Hee-Jin Ha, Hee Kim, Kwanghyun Choi, Young-Ho Kim, Hyerim Chang,
‡
∥
,‡
,†
Stephani Joy Y. Macalino, Jiyoun Lee, Sun Choi,* Jeewoo Lee*
†
Laboratory of Medicinal Chemistry, Research Institute of Pharmaceutical Sciences, College of
Pharmacy, Seoul National University, Seoul 08826, Republic of Korea
‡
National Leading Research Laboratory of Molecular Modeling & Drug Design, College of
Pharmacy and Graduate School of Pharmaceutical Sciences, Ewha Womans University, Seoul
0
3760, Republic of Korea
§
Medifron DBT, Sandanro 349, Danwon-Gu, Ansan-City, Gyeonggi-Do 15426, Republic of Korea
∥
⊥
Department of Global Medical Science, Sungshin University, Seoul 01133, Republic of Korea
Laboratory of Theoretical and Computational Biophysics & Faculty of Pharmacy, Ton Duc
Thang University, Ho Chi Minh City 75307, Vietnam
¶
Department of Pharmaceutical Chemistry, Hanoi University of Pharmacy, Hanoi 10000,
Vietnam
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
Alzheimer’s disease (AD) is an incurable, progressive neurodegenerative disease whose
pathogenesis cannot be defined by one single element but consist of various factors; thus, there is
a call for alternative approaches to tackle the multifaceted aspects of AD. Among the potential
alternative targets, we aim to focus on glutaminyl cyclase (QC), which reduces the toxic pyroform
of β-amyloid in the brains of AD patients. On the basis of a putative active conformation of the
prototype inhibitor 1, a series of N-substituted thiourea, urea and α-substituted amide derivatives
was developed. The structure-activity relationship analyses indicated that conformationally
restrained inhibitors demonstrated much improved QC inhibition in vitro compared to nonrestricted
analogs, and several selected compounds demonstrated desirable therapeutic activity in an AD
mouse model. The conformational analysis of a representative inhibitor indicated that the inhibitor
appeared to maintain the Z-E conformation in the active site, as is critical for its potent activity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Page 3 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
■
INTRODUCTION
Alzheimer’s disease (AD) is a progressive neurodegenerative disorder associated with
cognitive and behavioral dysfunction that ultimately leads to severe dementia. A variety of
therapeutic strategies have been developed to provide symptomatic relief; however, no disease-
modifying therapies have yet emerged because the causes and pathogenesis of AD have not been
completely elucidated. One of the leading theories of AD pathogenesis is “the amyloid hypothesis,”
which has now been put into question because clinical trials have largely failed to confirm it.1,2
Most of the tested drugs are small molecules and antibodies targeting Aβ producing enzymes and
Aβ aggregates, and they seemingly show little efficacy or unacceptable toxicity. Nevertheless,
there is a general consensus that anti-amyloid therapy is still a viable approach for early stage
patients, given that the accumulation of the Aβ peptides begins decades before any symptoms
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
appear. Moreover, there are numerous types of Aβ fibrils and oligomers that are either
neurotoxic4,5 or possibly neuroprotective ; thus, targeting specific forms of Aβ species may
6,7
provide effective treatment options in addition to other potential alternatives.
The N-truncated Aβ peptides generated from the N-terminal hydrolysis of Aβ are the main
constituents of amyloid plaques, represented as Aβn-40/42, where n ranges from 2 to 11.8-10
Interestingly, more than 50% of the total Aβ plaques in the AD brain comprise pyroglutamate Aβ
(pE-Aβ), AβΝ3pE3-40/42 and AβΝ3pE11-40/42, which are produced by cyclization of the N-terminal
glutamates (E), E3 and E11 in Aβ.11-13 The pE-Aβs appear to be present only in AD brain, not
normal brain, and aggregate quickly and irreversibly into soluble toxic oligomers that form inert
amorphous fibrils over a long period of time. In addition, pE-Aβs function as key initiators of AD
pathogenesis by acting as seeds for aggregation of Aβ1-40/42 and tau,14-16 suggesting that the pE-Aβs
may serve as a potential target to augment current anti-amyloid strategies.17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The pE-Aβ peptides are generated by glutaminyl cyclase (QC), which is mainly found in the
hippocampus and cortex of mammalian brain.18-20 Importantly, QC is overexpressed in the brains
of AD patients and animal models,21,22 and the knock-out of QC rescued cognitive function in AD
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
23
model mice. It has also been reported that small molecule inhibitors of QC efficiently reduced
2
4
22
the brain levels of pE-Aβ and Aβ plaques and improved memory deficits in AD mice. These
studies suggested that inhibition of brain QC activity may bring about beneficial effects for AD
patients.25,26 Several research groups have developed QC inhibitors that are based on imidazole
(
1)27 and benzimidazole core structures (2, 3)28,29 (Figure 1). The most prominent candidate is
PQ912 (3), developed by Probiodrug, which showed improvement in synaptic and neurological
functions in AD patients without significant toxicity in clinical trials.30,31 More recently,
Rubinsztein and colleagues introduced the N-methyltriazole-based inhibitor (4, SEN177), which
3
2
was discovered through 3D-pharmacophore model-based in silico methods. Similarly, Wu and
colleagues reported a diphenyl group conjugated imidazole inhibitor (5), which significantly
decreased the amount of pE-Aβ3-42 in a cell-based model and improved behavior in AD model
3
3
34,35
mice. Naturally occurring inhibitors
and other small molecule inhibitors developed by the
3
6
fragment-based approach have also been reported, demonstrating moderate in vitro activity. It
should be noted that while these QC inhibitors reduced brain levels of pE-Aβ3-42 and improved
cognitive functions, it is still in debate whether these inhibitors directly affected the QC activity in
brain. Brooks and colleagues reported that compound 1a did not appear to cross the blood brain
barrier (BBB) in mice, suggesting that the therapeutic effect of QC inhibition may be associated
with peripheral mechanisms.37
ACS Paragon Plus Environment

Page 5 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Reported QC inhibitors
Previously reported QC inhibitors are designed based on the N-terminal (Glu-Phe) structure
of its substrate Aβ3-42, containing the three principle pharmacophores including zinc-binding group
(A), hydrogen bonding donor (B) and phenyl group (C), respectively, in inhibitor 1. In addition to
this basic scaffold, our research group previously reported a series of QC inhibitors with a new
extended pharmacophore based on the N-terminal tripeptide (Glu-Phe-Arg) of Aβ3E-42. The
additional pharmacophore mimicking Arg (D) was tethered to the phenyl group of compound 6
38
(
Figure 1). The extended pharmacophore helped to improve the potency, ranging from 5 to 40-
fold enhancement compared to 1. In vivo experiments also indicated that compound 6 significantly
reduced the brain concentrations of pE-Aβ and total Aβ while restoring cognitive functions in an
AD animal model. Our molecular modeling studies revealed that the Arg mimetic 2-aminopyridine
group (colored in magenta) appeared to form strong interactions with the carboxylate group of
Glu327 in the QC binding site, supporting our hypothesis that the additional pharmacophore
provides an extra binding interaction. Since the 2-aminopyridine group in compound 6 can freely
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
rotate in the active site, there may be a favorable conformation that will generate more potent
3
9
inhibitors. The X-ray structure of the hQC-PBD150 (1a) complex showed that the ligand resided
in the hQC active site with a bent conformation, the so-called Z-E conformation. We performed
the docking analysis of the prototype inhibitor 1 using the same X-ray crystal structure of hQC,
and compound 1 in the similar Z-E orientation appeared to form favorable interactions in the active
site as shown in Figure 2.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. The docked pose of 1 in the Z-E conformation in hQC.
2+
1
is displayed as sticks with purple carbon atoms, and Zn is the magenta ball. The interacting
residues are shown as thin sticks with their carbon atoms in light blue. Hydrogen bonds are depicted
as black dashed lines.
Based on these observations, we hypothesized that a rigidification strategy to generate the
bent conformation of 1 might improve its binding potency. Therefore, we decided to incorporate a
conformational blocker to the thiourea/urea nitrogen that is proximal to the dimethoxyphenyl group
to induce the formation of Z-E conformers as described in Figure 3.
ACS Paragon Plus Environment

Page 7 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Strategy to facilitate Z-E conformation by N-substitution on thiourea/urea analogs
In this work, we synthesized a series of N-substituted thiourea/urea analogs of 1 and
investigated their structure-activity relationships (SAR) by performing in vitro enzyme activity
assays. We added alkyl groups of different sizes and added various aromatic and heteroaromatic
functional groups to search for additional binding interactions. Several potent inhibitors were
selected based on in vitro activity, and in vivo efficacy was further evaluated by measuring their
ability to reduce the formation of Aβ3(pE)-42 in acute and transgenic AD mice models. Lastly, we
performed molecular docking studies to identify the specific binding interactions of the most potent
compound and examined our hypothesis.
■
RESULTS AND DISCUSSION
Chemistry.
The N-substituted thiourea and urea compounds were obtained by the coupling reaction
between N-substituted 3,4-dimethoxyaniline fragments and the imidazole, 1-(3-azidopropyl)-5-
methyl-1H-imidazole or 3-(5-methyl-1H-imidazol-1-yl)propan-1-amine.
The syntheses of N-substituted 3,4-dimethoxyaniline derivatives were accomplished by
the two different synthetic pathways where either 3,4-dimethoxyaniline (7) or its 2-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
nitrobenzenesulfonamide (33) were directly N-substituted (Scheme 1). In the first pathway, most
of the N-substituted 3,4-dimethoxyaniline derivatives (8, 9 and 11-29) were synthesized by the
reductive amination of 7 using corresponding aldehydes or ketones, respectively (Method A). In
the case of N-cyclopropyl derivative 10, cyclopropanone was generated in situ from 1-ethoxy-1-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
0
(
trimethylsilyloxy)cyclopropane under reflux in acidic condition. The N-phenyl derivative 30 was
obtained by the Buchwald-Hartwig amination using iodobenzene (Method B). N-thiazol-2-yl
derivative 31 was prepared using 2-chlorothiazole through S Ar reaction under acidic condition
N
(Method C). N-oxazol-2-yl derivative 32 was prepared by the modified Robinson-Gabriel
4
1
method in which the phenyl carbamate of 7 was reacted with aminoacetaldehyde dimethyl acetal,
and then the resulting urea was cyclized in situ under acidic condition (Method D). In the second
pathway, N-substituted aniline derivatives were synthesized through 2-nitrobenzenesulfonamide
4
2
derivative 33, prepared by reacting 7 with 2-nitrobenzenesulfonyl chloride. The N-alkylation of
33 was conducted by either direct alkylation with corresponding alkyl halides for the syntheses of
3
4-35 or Mitsunobu reaction with the corresponding alcohols for the syntheses of 36-43. The
protected sulfonamide group was readily cleaved by a nucleophilic aromatic substitution with
thiophenol in basic media.
ACS Paragon Plus Environment

Page 9 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 1. Synthesis of N-substituted 3,4-dimethoxyaniline derivativesa
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: (a) (Method A) aldehyde or ketone, NaBH CN, AcOH, MeOH, 40 °C, 4 h; 1-
3
ethoxy-1-(trimethylsilyloxy)cyclopropane, AcOH/MeOH (1:1), NaBH CN, reflux, overnight; (Method B)
3
iodobenzene, Pd(OAc) , xantphos, Cs CO , dioxane, reflux, overnight for 30; (Method C) 2-chlorothiazole,
2
2
3
TsOH, EtOH, reflux, overnight for 31; (Method D) i) PhOCOCl, NEt then (CH O) CHCH NH , CH Cl ,
3
3
2
2
2
2
2
r.t., overnight; ii) c-HCl, MeOH, r.t., 2 h for 32; (b) 2-nitrobenzenesulfonyl chloride, NEt , CH Cl , 0 °C-
3
2
2
r.t., 4 h; (c) alkyl halide, Cs CO , DMF, r.t., 2 h; (d) hydroxyalkyl derivatives, Ph P, DEAD, CH Cl , r.t.,
2
3
3
2
2
overnight; (e) thiophenol, K CO , MeCN, r.t., overnight.
2
3
The syntheses of thiourea and urea compounds were accomplished by coupling N-
substituted 3,4-dimethoxyaniline (8-43) prepared in Scheme 1 with either one of the two imidazole
fragments (44, 45) (Scheme 2). 3-(5-Methyl-1H-imidazol-1-yl)propan-1-amine 44 was obtained
2
7
by following the previously described method, and its azide surrogate 45, 1-(3-azidopropyl)-5-
methyl-1H-imidazole, was prepared from 44 using the diazo-transfer reagent, imidazole-1-sulfonyl
43
44
azide hydrochloride, under CuSO catalysis in basic media. Aza-Wittig coupling reactions
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
between the synthesized N-substituted 3,4-dimethoxyaniline fragments and 45 under carbon
disulfide provided the N-substituted thiourea compounds 46-61. Meanwhile, the use of carbon
dioxide instead of carbon disulfide afforded N-substituted urea compounds (62-79, 81, 87, 88, 90-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
9
6 and 98) (Method E) and Boc-protected penultimate intermediates, which were hydrolyzed to
final amine compounds 80, 82-85 and 97 (Method F). N-Phenyl urea derivative 86 was obtained
by the coupling reaction using triphosgene between the two amines, 30 and 44 (Method G). N-
Piperidin-4-yl urea derivative 89 was obtained from N-benzylpiperidin-4-yl urea 94 by
hydrogenation in acidic condition (Method H).
Scheme 2. Synthesis of N-substituted thiourea and urea compoundsa
a
Reagents and conditions: (a) 1H-imidazole-1-sulfonyl azide hydrogen chloride, CuSO , K CO , MeOH,
4
2
3
r.t., overnight; (b) CS , Ph P, toluene, heat, 4 h to overnight; (c) (Method E) CO , Ph P, toluene, heat, 4 h
2
3
2
3
to overnight; (Method F) i) CO , Ph P, toluene, heat, 4 h to overnight, ii) CF CO H, CH Cl , r.t., overnight;
2
3
3
2
2
2
(Method G) triphosgene, pyridine, CH Cl , r.t. then 44, reflux; (Method H) i) CO , Ph P, toluene, heat, 4
2
2
2
3
h to overnight, ii) Pd/C, c-HCl, H , MeOH, r.t.
2
The syntheses of α-substituted amide derivatives were carried out by following the
pathway described in Scheme 3. 2-(3,4-Dimethoxyphenyl)acetonitrile 99 was alkylated with the
corresponding alkyl halides, which were then hydrolyzed to respective acids in basic aqueous
ACS Paragon Plus Environment

Page 11 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
medium at high temperature. The resulting acids were condensed with amine 44 using 1-ethyl-3-
(3-dimethylaminopropyl)carbodiimide (EDC) reagent to afford the final compounds, 100-108,
respectively.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 3. Synthesis of α-substituted amide compoundsa
a
o
Reagents and conditions: (a) NaH, alkyl halide, THF or DMF, r.t. to 60 C; (b) NaOH, H O, EtOH, reflux,
2
2
4 h; (c) 44, EDC.HCl, HOBt, NEt , r.t., overnight.
3
In vitro activity.
Structure Activity Relationships of QC inhibitors
The QC inhibitory activities of final compounds were determined by using a fluorogenic
substrate, Gln-AMC (L-glutamine 7-amido-4-methylcoumarin) and pyroglutamyl peptidase
(
pGAP) as an auxiliary enzyme, as described previously.38,45 These results are summarized in
Tables 1-4, together with the IC values of the previously reported inhibitors 1-5 for comparison.
50
First, we investigated the SAR of the N-substituted thiourea derivatives, as shown in
Table 1. The incorporation of alkyl groups, such as methyl (46), ethyl (47) and isopropyl (48)
groups, increased the potencies by 2-, 3- and 10-fold, respectively, compared to 1. As the size of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the conformation-blocking alkyl group increased, the QC inhibition also improved – likely due to
the formation of the Z-E conformer, the active conformation. Isobutyl (49) and cyclopentylmethyl
(50) derivatives also exhibited similar degrees of potency, suggesting that the improvement in
activity arises from the favorable conformational conversion, rather than from the increased
lipophilicity. When cyclic alkyl groups, including cyclopropyl (51), cyclobutyl (52), cyclopentyl
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(53), and cyclohexyl (54), were incorporated into the same position, the cyclopropyl derivative
(51) displayed a similar activity compared to 1, whereas cyclobutyl and cyclopentyl derivatives
exhibited improved potencies of up to 6-fold as the size of substituent increased progressively. The
incorporation of an aromatic ring generally yielded much improved potency, as demonstrated by
compounds 55-61 with IC values in the low nanomolar range. Notably, the N-4-fluorobenzyl
5
0
derivative (58) displayed excellent inhibition, with an IC value of 1.3 nM: much more potent than
5
0
its parent compound 1.
Table 1. IC values for inhibition of hQC by N-substituted thiourea compounds
5
0
a
Compound
R
IC (nM)
50
1
2
3
4
5
6
H
29.2
47.9
29.0
16.7
118
4.5
ACS Paragon Plus Environment

Page 13 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
4
6
7

Me
12.4 (±1.8)
8.8 (±1.2)

4
4
8
9


2.8 (±2.4)
6.8 (±2.7)
4.1 (±0.3)
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0

5
5
0
1


25.6 (±4.3)
8.6 (±0.8)
52
5
3

6.5 (±0.8)
4.7 (±0.4)
54


5
5
5
6
4.6 (±0.4)
3.4 (±0.9)


F
5
7
3.5 (±2.0)
Cl



5
5
8
9
F
1.3 (±0.8)
5.2 (±0.8)
Cl
60
5.4 (±3.7)
5.6 (±3.4)
N

6
1
N
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
The values indicate the mean of at least three experiments and the numbers in parentheses refer to the standard
deviation
Next, we investigated the SAR of the urea derivatives as shown in Tables 2-3. The
unsubstituted urea (62) was found to be much less potent than its thiourea surrogate 1. However,
we observed a similar trend in the thiourea derivatives once an alkyl group was introduced to the
urea nitrogen. The QC inhibition increases upon the addition of the N-alkyl groups with increasing
size as follows: acyclic (N-methyl (63) < N-ethyl (64) < N-isopropyl (65) < N-isobutyl (66) < N-
cyclopentylmethyl (67)) and cyclic alkyl substituents (N-cyclopropyl (68) < N-cyclobutyl (69) <
N-cyclopentyl (70) < N-cyclohexyl (71)). In particular, N-cyclopentylmethyl (67) and N-
cyclohexyl (71) derivatives exhibited 26- and 36-fold enhancements in potency compared to 62,
indicating that the effect of the conformational restriction was found to be more marked in the
series of urea than that of thiourea. Among the N-benzylic urea derivatives (72-76), the N-4-
fluorobenzyl derivative (75) also showed the most potent activity in this series, having the IC50
value of 3.9 nM. N-Pyridinyl derivatives (77-82) displayed a similar extent of inhibition with a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
range of IC = 8.7-32.5 nM.
50
Previously, we have found that the 2-aminopyridine moiety in 6 provided an additional
3
8
binding interaction with the enzyme. Therefore, we incorporated this group on the nitrogen with
various spacer lengths. Surprisingly, N-2-aminopyridinyl derivatives 84 and 85 with a 4-carbon
linker displayed excellent inhibitions, having the IC values of 3.6 and 1.6 nM, respectively.
5
0
Because both of these compounds contained a flexible linker, we believe that their potent activities
may arise from a strong ionic interaction between the 2-aminopyridyl group and the carboxylate of
Glu327 rather than conformational restriction based on our docking study (Figure S1). N-Phenyl
and N-heterocyclic derivatives in this series (86-88) showed only modest inhibitions.
ACS Paragon Plus Environment

Page 15 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. IC values for inhibition of hQC by N-alkyl/aryl substituted urea compounds
5
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Compound
62
R
IC (nM)
50
H
162 (±18.5)
78.7 (±20.6)
27.0 (±1.4)
6
6
3
4


Me
6
6
5
6


17.5 (±0.6)
7.6 (±3.4)

67
68
6.2 (±1.9)


38.5 (±16.3)
26.5 (±2.1)
6
7
9
0

19.4 (±5.1)
4.5 (±0.1)
7
1


72
73
13.3 (±1.3)
20.5 (±0.7)


F
7
4
12.9 (±6.5)
Cl
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6



7
7
5
6
F
3.9 (±0.5)
Cl
12.5 (±2.1)
8.7 (±1.4)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
77
N
N
N
N




7
7
8
8
9
0
F
30.6 (±13.3)
32.5 (±6.4)
Cl
NH2
23.5 (±9.1)
11.0 (±1.4)
81
82
N
NH2

17.6 (±7.3)
36.9 (±5.2)
3.6 (±1.1)
N
NH2


8
8
3
4
N
NH2
N
NH2

HN
8
5
1.6 (±0.3)
N
8
8
8
6
7
8


93.0 (±18.4)
69.6 (±11.4)
98.8 (±9.3)
N
O
N
S

ACS Paragon Plus Environment

Page 17 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
The values indicate the mean of at least three experiments and the numbers in parentheses refer to the standard
deviation
Based on the finding that the N-cyclohexyl urea derivative (71) was one of the most potent
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
inhibitors, we decided to study additional analogs containing bioisosteric piperidine surrogates, as
shown in Table 3. The N-piperidin-4-yl (89) and N-(1-methylpiperidin-4-yl) (90) derivatives
exhibited potent inhibitions, with IC values of 8.2 and 6.1 nM, respectively, comparable to that
5
0
of 71. Incorporation of bulkier groups, such as ethyl (91) and acetyl (92) groups, on the nitrogen
of piperidine led to marked reductions in potency. On the other hand, phenyl (93) and benzyl (94-
9
6) groups in the same position did not affect potency significantly. N-piperazinylethyl (97) and
morpholinylethyl (98) derivatives showed moderate inhibitions.
Table 3. IC values for inhibition of hQC by N-piperidinyl urea compounds
5
0
a
Compound
R
IC (nM)
50
8
9



NH
N
8.2 (±0.9)
6.1 (±1.5)
90
9
9
9
1
2
3
N
81.0 (±30.5)
53.9 (±5.2)
16.4 (±5.8)
O


N
N
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6


N
N
9
4
7.5 (±1.9)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
95
13.6 (±3.5)
F

N
9
6
23.3 (±6.4)
N


9
9
7
8
N
N
NH
15.0 (±3.9)
25.0 (±3.8)
O
a
The values indicate the mean of at least three experiments and the numbers in parentheses refer to the standard
deviation
Finally, we examined α-substituted amide derivatives as the bioisosteres of the N-alkyl
urea moiety. The unsubstituted amide (100) showed modest activity comparable to that of the urea
compound 62. However, incorporation of alkyl groups at the α-position led to a progressive
improvement in potency as conformational restriction was enhanced (100 < 101 (methyl) < 102
(
dimethyl) < 103 (cyclopropyl)). The ring expansion at the same position (104-106) did not
improve potency any further. α-Benzyl amide analogues (107, 108) showed only moderate
inhibition. In general, α-substituted amide analogs appeared to be less potent than N-substituted
urea analogs.
ACS Paragon Plus Environment

Page 19 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 4. IC values for inhibition of hQC by α-substituted amide compounds
5
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Compound
X
IC (nM)
50




1
1
00
01
183 (±12.7)
112 (±34.5)






1
1
02
03
77.3 (±18.5)
24.3 (±4.5)
1
04
37.1 (±12.2)
22.9 (±4.2)




105
106
77.5 (±0.7)
39.0 (±1.4)




1
1
07
08
81.6 (±18.9)
N
a
The values indicate the mean of at least three experiments and the numbers in parentheses refer to the standard
deviation
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Toxicity Study
On the basis of the in vitro QC inhibition, we selected the 24 most potent compounds
having IC values less than 10 nM for further evaluation. For in vitro toxicity study as shown in
5
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 5, we first examined the cell viability of each compound in a hippocampal neuronal cell line
HT-22) and found that a total of 11 compounds were nontoxic at 10 μM. We also examined the
(
human ether-a-go-go-related gene (hERG) potassium channel inhibition of each compound and
found that 16 compounds exhibited less than 50% inhibition at the same concentration.
Table 5. Toxicity studies of selected QC inhibitors
In vitro
IC50
(nM)
Cell viability
at 10 μM
(% of control)
hERG FP
(% inhibition
at 10 μM)
4
7
8.8
2.8
6.8
4.1
8.6
6.5
4.7
4.6
3.4
3.5
1.3
5.2
5.4
5.6
7.6
6.2
88.4
48
100
100
100
86.7
81.7
78.3
80.4
91.7
88.2
100
100
89.3
100
100
100
46.3
39.9
60.0
11.5
31.8
59.4
71.5
77.6
86.0
83.5
86.8
34.5
32.3
23.7
29.6
4
5
5
9
0
2
53
5
5
4
5
56
5
5
5
7
8
9
60
6
6
6
1
6
7
ACS Paragon Plus Environment

Page 21 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
7
1
4.5
3.9
8.7
3.6
1.6
8.2
6.1
7.5
88.6
12.2
40.6
33.2
48.0
40.8
26.3
46.1
73.5
75
90.7
100
100
82.2
88.3
100
92.4
7
8
8
7
4
5
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
89
9
9
0
4
In vivo activity.
Acute AD model
To assess in vivo efficacy, we tested all 24 compounds that were selected from our
in vitro activity study in the acute mouse model of AD. In this study, we injected 5 μg of human
Aβ3-40, together with each compound (25 mg/kg), into ICR mice (male, six weeks old) by
intracerebroventricular (icv) injection. We determined the activity of each compound on the next
day by measuring the levels of human AβN3pE-40 in the brain extracts of each of the mice. As shown
in Table 6, four compounds (77, 84, 89 and 90) suppressed greater than 20% of the formation of
AβN3pE-40 compared to the vehicle control. While the tested compounds mostly demonstrated
similar in vitro activities, their in vivo activities appear to vary. We believe that this apparent
discrepancy can be attributed to the use of different substrates and enzyme species for in vitro
assays (glutaminyl-AMC tested against human QC) and in vivo studies (human Aβ3-40 tested in
mice), as well as compound-specific characteristics such as solubility, enzyme reaction kinetics,
and metabolic rate. To examine whether these compounds exert their activity by penetrating the
blood-brain barrier (BBB), we again performed the same experiment by injecting the three effective
compounds (77, 84, and 90) via intraperitoneal (ip) injection while human Aβ3-40 was injected by
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
icv injection prior to the ip administration. As shown in Table 5, all three compounds exhibited
significant AβN3pE-40 lowering effects (> 20% inhibition) that were also comparable to the activities
observed in the icv administered experiments, suggesting that these compounds are likely to cross
the BBB. However, since these experiments are an indirect measure comparing biological effects
after the administration, possible peripheral association with QC inhibitory effect should be
considered for future mechanism studies.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 6. QC inhibition in acute model studies in vivo.a
%
inhibition of
In vitro
AβN3pE-40 formation
(icv injected) (ip injected)
NE
IC (nM)
5
0
4
7
8.8
48
2.8
6.8
4.1
8.6
6.5
4.7
4.6
3.4
1.3
3.5
5.2
5.4
5.6
7.6
6.2
4.5
3.9
6.0
NE
4
5
9
0
2.96
NE
52
5
5
5
3
4
5
1.17
4.01
15.28
6.66
NE
56
5
5
5
6
6
6
7
8
9
0
1
6
NE
0.73
5.31
11.58
NE
67
2.96
NE
7
7
1
5
NE
ACS Paragon Plus Environment

Page 23 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
7
7
8.7
3.6
1.6
8.2
6.1
7.5
25.98
22.87
22.74
84
24.38
13.70
26.26
35.46
19.40
8
8
9
5
9
0
NT
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
26.10
94
a
5
μL of human Aβ3−40 in PBS (1 μg/μL) was injected into the deep cortical/hippocampus of 5-week-old ICR mice
25 g, n = 4, male) using a stereotaxic frame to induce acute Aβ toxicity. Test compounds were administered via icv
or ip injection. Sandwich ELISA was performed for the quantification of the brain AβN3pE-40. NE: not effective, NT:
not tested
(
Additionally, we determined mouse and human liver microsomal stabilities of these three
compounds. The remaining percentage values of intact compounds after 30 min for 77, 84 and 90
were found to be 12.8%, 21.6% and 89% for mouse and 44%, 43.1% and 100% for human,
respectively, indicating that 90 was the most stable in metabolism. We further tested compound 90
for selectivity against isoQC, an isozyme of QC with a differential substrate specificity.46
Compound 90 was found to have an IC value of 304 nM for isoQC, which was more than 50-fold
5
0
greater than that for QC, indicating that 90 is highly specific for QC. We also determined the IC50
value of 90 in mouse QC as 28.4 nM, indicating that 90 is an hQC selective inhibitor.
AD Model
To assess the desired therapeutic effects of 90 in AD, we performed long-term in vivo
studies in 5XFAD transgenic mice. It has been reported that 5XFAD mice generate severe AD
4
7
related pathologies with massive increases of Aβ and pyroform Aβ in the brain ; this model is thus
suitable to study Aβ-induced neurodegeneration. Compound 90 was administered by ip injection
to 32-week-old 5XFAD mice at a dose of 25 mg/kg every day for 4 weeks. After finishing the 4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
week administration, the brain concentrations of AβN3pE-40/42 and Aβ40/42 were measured by
sandwich ELISA. As shown in Figure 4, compound 90 significantly reduced the brain
concentrations of AβN3pE-40 (Figure 4A) and AβN3pE-42 (Figure 4B) by 57.9 and 50.5 percent,
respectively, compared to the vehicle control. Compound 90 also reduced the Aβ and Aβ levels
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
0
42
(Figure 4C and 4D) substantially, by 36.6 and 35.1 percent without significant toxicity, indicating
that the inhibition of QC not only reduced the amount of pE-Aβ in brains but also contributed to
the overall reduction in Aβ levels as previously reported.48,49
Figure 4. Inhibition of the formation of human AβN3pE-40/42 and Aβ1-40/42 in 5XFAD mice.
(a) Brain concentrations of AβN3pE-40; (b) brain concentrations of AβN3pE-42; (c) brain concentrations
of Aβ ; (c) brain concentrations of Aβ . Data were analyzed by unpaired t-test or one-way
4
0
42
ANOVA. (*: p < 0.05, **: p < 0.01).
ACS Paragon Plus Environment

Page 25 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Computational study.
Our initial hypothesis of this study is that since the bent conformation of compound 1 with
a Z-E orientation is the most active conformation, the N-substituted thiourea/urea analogs with a
conformational blocker will favor the Z-E conformation and thus improve the overall potency.
Both in vitro and in vivo activity tests showed enhanced activity compared to the unsubstituted
analogs, which is consistent with our hypothesis. To better understand how our conformational
restriction strategy worked to improve QC inhibition, we performed comprehensive computational
analysis.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We first performed conformational analysis for the prototype inhibitors 1 and 62. The
unsubstituted thiourea 1 displayed the Z-E form as the dominant conformation, whereas the
unsubstituted urea 62 had both Z-Z and Z-E conformations, with Z-Z as the major form. We also
employed an incremental conformation search by calculating the dihedral angles of the
urea/thiourea groups. The plot showing the relationship between energy and conformation revealed
that lower energy was observed for 62 when its urea group was in the Z-Z form, suggesting that 62
was more stable in the Z-Z conformation (Figure 5A). On the other hand, the corresponding
thiourea group of 1 was more stable in the Z-E conformation (Figure 5B). Given that the bioactive
conformation of the QC inhibitor favors the Z-E form and that 62 has a mixture of Z-Z/Z-E
conformations, this analysis explains why compound 62 displays 5 times less inhibitory activity
than that of 1.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Plots of energy versus incremental conformational changes from calculated
dihedral angles
(A) Thiourea in 1, (B) urea in 62, and (C) N-piperidinyl urea in 90. The X axis represents the
dihedral angle (C-C-N-C) found near the 5-methyl imidazole ring, while the Y axis represents the
dihedral angle (N-C-N-C) near the dimethoxy phenyl group. Dihedral angle degree close to 0
denotes the E-form, and close to 180/-180 denotes the Z-form. The plots show that low energy
conformations occupied mostly Z-E form in thiourea 1 and N-substituted urea 90. In contrast,
diversity (both Z-Z and Z-E) of low energy conformations exist for the urea compound, 62.
Using the most stable conformers from the conformational analysis, we optimized and
calculated the total free energy of both 1 (Z-E) and 62 (Z-Z). Only minimal changes were observed
after optimization, as both structures maintained the stable conformations. The resulting total free
energies of 1 and 62 are -1390.33 and -1067.37 kcal/mol, respectively. These values indicate that
the structure of compound 1 is more stable than that of 62, whose higher energy may result from
the unconstrained dihedral of its urea group leading to a mixed conformational population. Among
the synthesized compounds in which different substituents were added to the proximal nitrogen of
urea/thiourea, 90 exhibited excellent inhibitory activity toward hQC, with 5- and 26-fold increased
activity compared to 1 and 62, respectively. The conformational analysis of 90 demonstrated that
the Z-E form was its sole conformation with high stability, consistent with our design strategy
(Figure 5C). The most stable conformer of 90 has a total free energy of -1357.68 kcal/mol, which
ACS Paragon Plus Environment

Page 27 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
is comparable to that of 1. This analysis supported our hypothesis that the design strategy restricted
to the bioactive Z-E form provided more potent inhibitors against hQC.
To investigate the binding interactions of 90, the protonated piperidine form of 90 at pH 7.4
was used for Glide QM-Polarized Ligand Docking. The results indicated that 90 was docked very
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
9
well at the active site of hQC with Z-E conformation as anticipated. The binding interactions
examined were as follows: the 5-methyl imidazole group chelated with zinc and formed H-bonding
interaction with the indole NH of Trp329. Additionally, it produced hydrophobic interactions with
Leu249 and Trp207. The oxygen of urea formed an H-bonding interaction with Gln304. The 3,4-
dimethoxy phenyl group displayed hydrophobic interactions with Ile321, Pro324, and Phe325, in
which the phenyl ring presented π-π interactions with Phe325. The N-substituted piperidinyl group
not only restricted the conformation but also showed hydrophobic interactions with Tyr299, Ile303,
and Val302 (Figure 6).
Figure 6. Docked result of 90 in hQC. (A) Binding interactions of 90 at the active site of hQC.
2
+
90 is displayed as sticks with cyan carbon atoms, and Zn is the purple ball. The interacting
residues are shown as thin sticks with their carbon atoms in light blue. Hydrogen bonds are depicted
as black dashed lines. (B) 2D representation of the interactions of 90 with the active site residues
of the hQC. Hydrophobic interactions are marked in light brown. Hydrogen bonds are shown as
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
red-dotted arrows with directionality. The π-π stacking interaction is marked by a blue disc and
arrow.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
■
CONCLUSION
In this work, we developed a series of N-substituted thiourea and urea surrogates of 1 as
conformationally restricted analogs that favor a bent conformation (Z-E conformation). Our in-
depth SAR analyses indicated that conformational restriction by the proximal N-substituent was
able to improve the potency with respect to QC inhibition by up to 20-fold for thiourea, 100-fold
for urea and 8-fold for amide. Among the library of compounds, N-4-fluorobenzyl thiourea (58)
and urea (75) analogs exhibited excellent QC inhibition, with the IC values of 1.3 and 3.9 nM,
5
0
respectively. For further evaluation of the selected inhibitors with IC values of less than 10 nM,
5
0
we measured the levels of AβN3pE-40 in an acute model and examined their cytotoxicities, hERG
inhibitions and metabolic stabilities. Among the selected inhibitors, 90 showed the most promising
efficacy and druggable profile and thus was evaluated in the AD model for in vivo efficacy. After
4
weeks of administration, 90 significantly reduced the brain concentrations of pyroform Aβ and
total Aβ in the 5XFAD mouse model. The conformational analysis of the two prototype inhibitors
and 62 indicated that unsubstituted thiourea 1 displayed mostly a bioactive Z-E conformation,
1
whereas unsubstituted urea 62 existed in a mixed conformation with a major Z-Z form, explaining
why 62 is much less potent than 1. The docking analysis of 90 indicated that the Z-E conformer is
the dominant form and displayed the three key interactions in the active site, supporting our
hypothesis that the introduction of a conformational blocker favors the active conformation.
Although recent clinical failures in anti-amyloid therapies might have disappointed researchers in
the related field, amyloid-targeting drugs can still be beneficial for early-stage AD patients, and
7
they offer advantages for combination therapy by lowering the levels of toxic Aβ species. We
ACS Paragon Plus Environment

Page 29 of 63
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
demonstrated that highly potent and selective QC inhibitors with promising in vitro and in vivo
activity can be rationally designed, which may open a new avenue for alternative therapeutic
strategies.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
■
EXPERIMENTAL SECTION
General. All chemical reagents were commercially available. A Buchi B-540 apparatus was used
to determine melting points, which are uncorrected. Silica gel flash column chromatography was
performed on silica gel 60, 230–400 mesh, Merck. The PLC plates used in the present study were
PLC silica gel 60 F , 1 mm, Merck. Proton NMR spectra were detected on a JEOL JNM-LA 300
2
54
at 300 MHz, Bruker Analytik, DE/AVANCE Digital 400 at 400 MHz, a Bruker Analytik,
DE/AVANCE Digital 500 at 500 MHz, and a JEOL JNM-ECA-600 at 600 MHz. Chemical shifts
are calculated in ppm units, with Me Si as a reference standard. A VG Trio-2 GC−MS instrument
4
and a 6460 Triple Quad LC−MS instrument were used to determine mass spectra. All final
compounds were purified to > 95% purity using high-performance liquid chromatography (HPLC)
on an Agilent 1120 Compact LC (G4288A) instrument using Agilent TC-C18 column (4.6 mm ×
2
50 mm, 5 μm) and an Agilent Eclipse Plus C18 column (4.6 mm × 250 mm, 5 μm).
General procedure for reductive amination (Procedure A). To a mixture of 3,4-
dimethoxyaniline 7 (1.0 eq) and ketone or aldehyde (1.0 eq) in MeOH was added acetic acid (5.0
eq), followed by NaBH CN (3.0 eq). After stirring at 40 °C for 4 h, the reaction mixture was
3
basified by NaHCO solution and then extracted with CH Cl . The organic layer was concentrated
3
2
2
in vacuo and purified by silica gel column chromatography (EtOAc/ n-hexane, 1:4 to 1:1) to obtain
the secondary amine 8-29.
General procedure for N-alkylation (Procedure B). A mixture of sulfonamide 33 (1.0 eq), alkyl
ACS Paragon Plus Environment